This method reformulates the chance-constrained problem as a tractable problem by creating a two-stage stochastic programming formulation. The first stage is selecting the best crude oil combination among several candidates and their price amounts to maximize expected gross margin across all scenarios. In the second stage, the uncertainties are realized and the optimal operations of the plant are implemented such that the market demand and quality specifications are satisfied. In order to obtain the global optimal solution for this problem, the variable discretization, feasibility and optimality-based domain reduction techniques are integrated into the non-convex generalized Benders decomposition (NGDB) methodology. The enhanced NGBD approach is able to find and verify a global optimal solution within a couple of hours compared to state-of-the-art commercial software which takes several days.